KR100216123B1 - Monitoring system of water quality - Google Patents

Abstract

According to the present invention, by reducing the amount of information to be processed, it is possible to monitor abnormal behavior of fish without using expensive equipment, and to minimize errors in judgment.

Technical composition of the present invention; Image acquisition means for extracting the behavior of the fish from the tank; Information extraction means for analyzing the behavior of the fish extracted from the image acquisition means and extracting each feature information of the fish; Memory means for embedding information according to a fish behavior pattern; Model matching means for matching feature information extracted from said information extracting means with information according to a behavior pattern of a fish embedded in said memory means; Analysis means for analyzing the matched signal of the model matching means; Pollution degree calculation means for calculating a pollution degree by the result of the analysis means; Water quality abnormality monitoring system comprising an alarm device for transmitting an alarm signal when the hook of the pollution degree calculation means exceeds a predetermined value.

Description

Water Abnormality Monitoring System

1 is a schematic diagram of an overall system according to the present invention.

2 shows a configuration of the control device of the present invention.

3 is a view showing the operation of the control device of the present invention.

4 is a diagram illustrating a process of extracting an image and removing noise in FIG. 3 in detail.

FIG. 5 is a diagram illustrating a process of processing each feature information in FIG. 3 in detail.

(Industrial response field)

In the water treatment plant, in order to monitor whether or not toxic substances are mixed in the raw water, a part of the raw water is described as fish, for example, fish such as a fish, a crucian carp, a goldfish (hereinafter, collectively referred to as 'fish'). A system for drawing fish into a tank in which fish are raised and monitoring the behavior of fish is known from Japanese Patent Laid-Open Nos. 63-307385, 63-179252, 52-83663 and 57-1979. .

This kind of monitoring system monitors the inflow of toxic substances into raw water by using phenomena such as fish behaving strangely or dying when toxic substances are mixed in raw water. Continuously monitoring the behavior and image processing to determine the condition of the fish to detect the inflow of toxic substances in the water treatment plant or sewage treatment plant.

Therefore, such a system basically processes a tank for raising fish, a photographing value for photographing the behavior of the fish in the tank, and outputs image information, and an image having a different photographing time from the image information obtained from the photographing apparatus. And state means output means for outputting a state amount of fish, and means for detecting and displaying whether or not the state amount of fish has exceeded the set level.

Japanese Patent Laid-Open No. 2-161352 discloses a system for monitoring abnormal water quality by monitoring the moving speed of fish in a tank. In this system, the behavior of the fish in the tank is monitored by the photographing apparatus, converted into an omnidirectional signal, and output, and the signal output from the photographing apparatus is transmitted to the moving speed detecting means. The moving speed detecting means obtains the average moving speed of the fish based on the image of the fish obtained by the imaging device, and then adds it to the moving speed determiner. The transport islets determiner compares the first moving speed and the second moving speed with the average speed, and determines that abnormal behavior is caused by the inflow of toxic substances when the fish shows more active movement than the first moving speed. The movement less than the speed indicates the death of the fish due to the inflow of toxic substances, and the result is the ON / OFF signal of the alarm device to sound the alarm.

However, these systems have to keep track of their movements in order to observe the abnormal behavior of the fish, so there is a drawback that expensive equipment is required due to the increased amount of information to be processed.

Accordingly, an object of the present invention is to provide a system capable of monitoring abnormal behavior of fish without using expensive equipment by reducing the amount of information to be processed.

Another object of the present invention is to provide a system that can monitor abnormal behavior of fish and minimize errors in determination without using expensive equipment by reducing the amount of information to be processed.

The present invention for achieving the object, the fish tank containing the fish, the image acquisition means for extracting the behavior of the fish from the tank, by analyzing the behavior of the fish extracted from the image acquisition means to analyze each feature information of the fish A model for matching the information extraction means for extracting, the memory means in which the information according to the behavior pattern of the fish is embedded, the characteristic information extracted from the information extraction means, and the information according to the behavior pattern of the fish embedded in the memory means Matching means, analysis means for analyzing the matched signal of the model matching means, pollution degree calculating means for calculating the pollution degree by the results of the analysis means, an alarm signal when the result of the pollution degree calculating means exceeds a predetermined value It consists of a sending alarm device.

The image acquiring means includes image means for converting fish behavior from the tank into an image signal and image signal converting means for converting the converted image signal into a digital signal. The image means may be a color CCD camera or the like. . The image signal converting means converts the analog signal acquired by the color CCD camera into a digital signal that can be recognized by the signal processing apparatus, that is, the signal processing apparatus can sample and quantize the analog one-dimensional signal to recognize the signal. It converts into discrete signal. The meteor board is used as the video signal converting means.

The information on the behavior pattern of the fish, that is, the information produced by the learned model, is built in the storage means in advance. As such storage means, a storage device such as a magnetic disk or EEPRON may be used.

The behavior of fish in water containing toxic substances may exhibit different behavior patterns. In the present invention, the feature information extracted by the behavior of fish includes floating, avoidance, clustering, diffusivity, and mobility. Used as information to estimate pollution.

Therefore, in order to extract the feature information according to the behavior pattern of the fish, the information extracting means processes each image from the image acquisition means in units of pixels, and the brightness information and the color information contained therein are extracted as the primary information as the information. .

The extracted first information is transmitted to the matching means to extract the area of the object.

The region of the object includes object extraction means for extracting an object region by determining whether all images included in each pixel are points belonging to the object or a point corresponding to a background, because all images are processed as pixels. In this case, it is determined by the size of each point included in the pixel whether it is a point corresponding to the object, a point corresponding to the background, or noise.

The information extracted by the matching means is applied to the analysis means, and the operability of the fish is questioned. The motion analysis of fish is performed by the information about the image of the current time and the images of the previous time. In this way, the secondary information, that is, floating bias, mobility, clustering, diffusivity, and the like, is extracted, and the secondary information is applied to the calculation means and used to determine the degree of contamination.

The result of the pollution benefit calculation means is compared with a predetermined value by the remark means and used as a signal to operate the alarm which sends an alarm signal as a result.

Hereinafter, a device according to the present invention will be described with reference to the accompanying drawings.

FIG. 1 is a schematic diagram of the whole system according to the present invention, and FIG. 2 shows the configuration of the control device of the present invention. 3 is a program showing the operation of the control device of the present invention. FIG. 4 is a program detailing a process of extracting an image and removing noise in FIG. 3. 5 is a program showing in detail the process of processing each feature information in FIG.

Referring to FIG. 1 which shows the whole system 1, fish are raised in the tank 12. As shown in FIG. The behavior of the fish raised in the tank 12 is converted into an image signal by the CCD camera 14 and sent to the image acquisition unit 18 through the cable 16. The image acquirer 18 samples, quantizes, and converts the analog signal from the CCD camera 14 so that the signal processing apparatus can recognize the digital signal. This digital signal is applied to and processed by the signal processing apparatus, and the signal processing apparatus 20 controls the system by the central processing unit 181 and the central processing unit 181 controlling the entire system 1 according to the present invention. It consists of a ROM 184 in which necessary programs are embedded, and such a signal processing apparatus may be a general personal computer, that is, a PC. At this time, the information according to the behavior pattern of the fish, that is, the information generated by the learned model, may be embedded in a magnetic disk which is an external memory device, for example, a floppy disk or a hard disk which is a fixed memory.

Moreover, the use of such external memory stores the information generated by the learning model according to the environment by using a memory element such as EEPRON, which can be fixed as a system to solve the disadvantage of inflexibility that cannot be adapted to various environmental changes. And lower the construction price of the entire system.

On the other hand, the central processing unit 181 of the signal processing unit is connected to the image acquisition unit 18 through the interface I (183) connected via the bus 182. In addition, the central processing unit 181 is connected to the alarm unit via the interface unit II (188) connected to the bus 182.

The operation of this configuration of the present invention will be described with the program according to FIG.

When power is applied to the entire system 1, the central processing unit 181 samples the analog signal from the CCD camera 14 monitoring the behavior of the fish being bred in the tank 12 (step 201). Quantization and conversion into digital signals (steps 202 and 203) are then processed into two-dimensional information on a matrix in pixels, which are transmitted to the central processing unit (step 204). Then, a point less than a predetermined size among the two-dimensional circle information on the pixel-by-pixel matrix is treated as 0 to eliminate noise (step 205), and an object is extracted (step 206). The floating object is determined in this way, and the result is stored in the RAM (step 207). The process of determining the richness can be configured as one subroutine, and the process of determining the richness will be described in detail below.

After the determination of the floatation, the bias is determined for the extracted object and the result is stored in the RAM (step 208). The process of determining the bias may be configured as one subroutine, and the process of determining the bias is described in detail below.

After the determination of the bias is completed, the mobility is determined for the extracted object, and the result is stored in the RAN (step 209). The process of determining the mobility may be configured as one subroutine, and the process of determining the mobility will be described in detail below.

After the determination of the mobility, the clustering of the extracted object is determined and the result is stored in the RAM (step 2110). The process of determining such clustering may be configured as one subroutine, and the process of determining clustering will be described in detail below.

When the determination on the clustering is completed, the biasedness of the extracted object is judged and the result is stored in the RAM (step 211). The process of determining the diffusivity may be configured by one subroutine, which is described in detail below.

The central processing unit performs the parity reasoning based on the determination result of the secondary information such as floatingness, bias, mobility, clustering, and spreadability stored in each RAM (step 212). Then, it is determined whether or not the result of the fuzzy inference exceeds the set threshold (step 213). If the threshold value is higher than the threshold (step 214), if it is less than the threshold value, the controller proceeds to step 201 to monitor the fish in the tank continuously. do.

[Area Extraction]

In this system using the vision-base approach, images must be recognized to measure the behavioral patterns of fish in the tank. For this purpose, a process of recognizing the position and size of the fish from the image of the tank obtained from the CCD camera of FIG. 4 is included.

Divide each pixel of the two-dimensional image obtained by the CCD camera into pixels arranged in a matrix, and perform a binarization operation for each pixel into a fish area and a background area. If it belongs to the area, it has zero.

This process binarizes the image using the threshold value TH calculated in the learning process. If r is greater than TH, 1 is arranged, and if r is less than TH, 0 is mapped to each pixel in a matrix (step). 301,302,303).

This process is examined for every pixel (step 304). The labeling operation is performed to have the result of the binarization operation in steps 302 and 303 having information on connectivity between pixels in a matrix array of 0s and 1s corresponding to each pixel (step 308).

This labeling operation is intended to allow binarized two-dimensional information to have information about the location and size of fish because it does not have information about connectivity between pixels. Therefore, the labeling operation calculates the connectivity between pixels. In this system, one-pass labeling operation is performed.

First, labels for pixels (x, y) are performed by comparison with adjacent pixels. In other words, it is determined whether adjacent pixels are labeled (step 308), and if so, the process proceeds to step 309 to make the pixel value equal to the adjacent value or to set the threshold value (steps 309 and 410). Proceed and come to adjacent pixels. This process is reviewed for every pixel by staff 312.

After reviewing all the pixels, it is determined whether the labeled size is larger or smaller than the predetermined size (step 313). If smaller than the predetermined size, it is determined as noise and removed (step 314). This will label all the pixels.

Each labeled pixel is regarded as an object and the hungry pattern of the fish is analyzed.

The behavior patterns of fish in the tank are classified into four stages.

The first step is to acquire an image from the tank by the image acquisition device.

In order to recognize the position of the fish in the tank, the system uses a vision-based approach to acquire the image from the tank in real time and transmit the two-dimensional signal to the system.

The region thus obtained is extracted through the recognition process. That is, in the recognition process, each pixel is classified into an object and a background part according to pixel values in the acquired image to extract information such as the position and size of the fish. Each feature information extracted by the actions of the fish, namely, Floatability, avoidance, clustering, and diffusivity are extracted.

The pollution calculation module determines whether the fish behavior pattern is normal using the feature values calculated in the feature value calculation models, and determines the pollution level of the water based on the result.

The system, using a vision-based approach, must recognize images to measure the behavioral patterns of fish in the tank. That is, behavior patterns are recognized through the process of recognizing the position and size of the fish in the image of the tank acquired by the CCD camera.

In order to simplify the amount of computation, a binarization operation is performed to classify each pixel of the two-dimensional image obtained by the CCD camera into a fish area and a background area. Each pixel of the image is classified into two groups through a binarization operation, and has a value of 0 and 1. In other words, if it belongs to the fish area, it has a value of 1, and if it belongs to the background area, it has a value of 0.

The binarization operation does not have binary two-dimensional information, that is, information about connectivity between pixels, and cannot calculate information about the position or size of the fish.

The labeling operation is an operation for calculating connectivity between pixels. In this system, one-pass labeling operation is performed.

Next, using the information on the size of the area calculated during the labeling operation, a noise removal operation is performed to remove the area that is incorrectly recognized as a fish area during the binarization operation. In other words, an area whose size is too small is regarded as noise and removed from the fish area.

[Feature Extraction Process]

The feature extraction module extracts five predefined feature values. These values are precalculated using information on the size and location of the fish obtained from the current and binary images.

The feature extraction module calculates five feature values to measure the behavioral patterns of fish. These feature values are defined as the features that can best classify fish behavior patterns in normal water quality and fish behavior patterns in abnormal water quality. Therefore, it is efficient for these feature values to use floating, avoidance, diffusivity, and clustering properties.

In this model, the feature value calculated above plays a role in determining whether the fish behavior pattern is abnormal. In other words, fuzzy inference is performed using the input feature values and the fuzzy rules stored in the knowledge base to determine the pollution level of water.

Fuzzy rules and membership functions stored in the knowledge base can be modified by the user to improve system performance. Each control variable may have two abnormal and normal values, and there are five control variables corresponding to each characteristic value.

The information extraction means extracts information according to the behavior pattern of the fish. Each image from the image acquisition means is processed in units of pixels so that brightness information and color information included therein are extracted as information. The extracted information is transmitted to the matching means to extract the area of the object. The area of an object is determined by the size of each point included in a pixel whether all images are processed as pixels and whether each point included in each pixel belongs to an object, a background point, or noise. do.

The information extracted by the matching means is applied to the analysis means to analyze the operability of the fish. The motion analysis of fish is performed by the information about the image of the current time and the image of the previous time. The information extracted in this way is suspended, unbiased, mobile, clustered, diffused, and the like. The information is applied to the calculation means and used to determine the degree of contamination.

4 shows a prochart for the process of extracting features.

First, vertical and soup profiles are created to calculate floatability and avoidance (step 401). Then, using the profile calculated from the current image and the previous image to calculate the floating and avoidance and store the result in the RAM (step 402).

Floatability is a measure of the degree to which fish float to the surface. It is defined to have a value close to 1 if many fish are concentrated near the surface, and to 0 if a fish is located at the bottom of the tank.

Avoidance is a measure of the response of fish to avoiding contaminants. That is, since water is introduced through the inlet located on the left side of the tank, the pollution degree of the left side of the tank is higher than that of the right side. Thus, fish respond to the less contaminated side, and avoidance is a measure of the fish's response.

Next, the cross value is calculated from the acquired image (step 403), the clustering of the current image is calculated using the cross values calculated from the current and previous images, and the result is stored in the RAM (step 404). Then, the diffusivity of the current image is calculated using the calculated value and the result is stored in the RAM (step 405).

Clustering is a measure of the degree of clustering of fish. It is defined to have a value close to 1 when a large amount of fish is concentrated, and a value close to 0 when evenly distributed throughout the cutlery.

Clustering is a measure of the degree of clustering of fish. It is defined to have a value close to 1 when there are many fish, and close to 0 when distributed evenly throughout the tank.

Diffusion measures the extent to which fishes move in or out relative to the center of the fish community. That is, it has the same effect as the first derivative of Close.

Next, the shape vector is calculated from the acquired image (step 406), and the mobility of the fish is approximately calculated using the shape vector calculated from the current image and the previous image, and the result is stored in the RAM (steps 407 and 408).

The speed of fish is calculated using the speed of individual fish. In this case, the system must track each fish internally, so this method of performing a large amount of calculation is not suitable in the present system which performs the real-time processing. Therefore, the approximate method is used to measure the movement speed of fish. The mobility is approximated by the shape vector.

Fuzzy reasoning is performed with each result calculated in this way (step 212 in FIG. 3), and an alarm signal is sent when the result of the pollution degree calculating means exceeds a predetermined value (steps 213 and 214 in FIG. 3).

For fuzzy inference, we first calculate the degree of matching of each fuzzy rule according to the input values. With this degree of agreement, we determine how much the conclusions of each rule affect. The conclusions of each rule produce an overall conclusion on pollution levels. Convert the result of the selected fuzzy inference into a value that can be used externally. And if the result inferred to pollution degree is over a certain value, it sends out a warning signal.

The present invention monitors water accidents caused by inflow of unspecified pollutants, such as rivers, lakes, treatment plant discharge waters, and general water sources, especially those that have a direct relevance to life. This can minimize the economic and sanitary damages to the post-processing process and water users. By using the above configuration, the amount of information can be reduced and general personal computers can be used without using expensive shopping bags. A system can be provided to monitor abnormal behavior of fish.

Claims (3)

A tank containing the fish, and image acquisition means for extracting the behavior of the fish from the tank; Information extraction means for analyzing the behavior of the fish extracted from the image acquisition means and extracting each feature information of the fish; Memory means for embedding information according to a fish behavior pattern; Model matching means for matching feature information extracted from said information extracting means with information according to a behavior pattern of a fish embedded in said memory means; Analysis means for analyzing the matched signal of the model matching means; Pollution degree calculation means for calculating a pollution degree by the result of the analysis means; And an alarm device that raises an alarm signal when a result of the pollution degree calculating means exceeds a predetermined value. The water quality monitoring system according to claim 1, wherein the image acquisition means comprises image means for converting fish behavior from the tank into an image signal and image signal conversion means for converting the converted image signal. 3. The image processing apparatus according to claim 2, wherein the image means is a color CCD camera, and the image signal converting means converts an analog image signal acquired by the color CCD camera into a digital signal that can be recognized by a signal processing apparatus. Water quality monitoring system.

Images